The Viable Systems Model developed by Stafford Beer in the 1970's suggests that a fractal approach to organisational design is capable of creating organisations that can respond more fully and far more rapidly to customer markets than current hierarchical models.
The VSM provides a template for analysis of existing organisations that helps to identify where real problems are occurring. Composed of 5 systems, system 2 acts very much like the spinal cord in the human being (or should) in that selected required information is passed from millions of external environment sensors up the chain of systems and messages from levels 3,4 & 5 are passed back to the "operational" level responsible for the major transformation activities which deliver an exchange of value with the external environment.
In practice System 2 is frequently hardly developed at all and sometimes simply does not exist.
In VSM terms this is because a hierarchical organisation is not capable of analysing and making use of the abundance of data available across the system's touch points with the external environment. All decisions affecting strategy and development are taken at the very limited senior levels where there is simply insufficient capability to process all the data available (if it is even allowed to impinge on the higher systems).
By contrast VSM proposes that each level in the organisation has a strategy and development function which are syndicated into a whole via system 2. This means that a balance is achieved between systems 4 and 5 and their fractal correspondents in system 1.
In other words system 1 is also a VSM in it's own right coordinated with the other components of the system level above and below it by system 2.
Problems with making an effective system 2 a reality are usually along the lines of "I don't want my operations people spending time on anything but operations" This would be a standard hierarchical view in for example a business which was based on a fee-earning model. In this case the problem is particularly acute because hours not on fees translate straight to lost production.
How do you gather the inputs from tens of thousands of people and process them so that they make any sense overall and how do you actually turn it into the basis for action?
Interestingly VSM has languished pretty much for the last 40 years except for those whose sense of dissatisfaction with the received models of organisational structure drove them to seek something that would work better. I think the problem may well have been the difficulty in providing scale of communications and collaborative environments that could facilitate a really effective system 2.
However now we have a number of technologies emerging from the development of the World Wide Web that make communication much easier. In fact it is so easy that it is a challenge to deal with the variety of inputs. In amongst a number of technologies loosely bundled under the title of "Enterprise 2.0" are prediction markets which have a very wide range of potential uses that are only beginning to be explored. One however which may be of crucial importance, especially to large enterprises is the way in which employees in a company can be encouraged to participate in a metaphorical trading game that uses self selection through the buying of "stocks" in the value or likely outcome of questions listed on the market.
Through a mechanism of price discovery a consensus of a very large number of people can be developed very rapidly using up very little of their time. The design and implementation of the market however is crucial.
It is here that I think the VSM and prediction markets meet in a potentially hugely fruitful union. If you attempt to drive the prediction market top down setting the agenda for all questions you will lose most of the value. If on the other hand you allow anyone to include any question at all you will soon have an unmanageable chaos. Ashby's Law of Requisite Variety again comes into play.
However VSM has an answer. The System 2 structure (of which I suggest a Prediction Market can form a major part) can itself be configured as a VSM with multiple layers. This gives us a key component for seeing how Prediction Market technology can be applied in a fractal manner with multiple markets converging to provide essential inputs to systems 3,4 & 5 as well as providing a means for local strategy and development to be aligned with corporate System 5 policy making.
So far so good (I think). But all of this requires the development of an approach to the organisation which is an appropriate balance between autonomy and central control.
I hope to address those vital and interesting questions a little bit in the future.
Best Wishes
Mike
